Our objective is to understand the function of intermediate filament (IF) proteins and their differential expression in tissues. The epidermis has been chosen as a model system, since it is possible to (l) culture human epidermal cells under conditions where differentiative properties are retained; (2) use keratinocyte-specific promoters to target expression of foreign genes to the epidermis of transgenic mice; and (3) biopsy skin from transgenic or embryonic stem (ES) cell-derived mice, enabling extensive analyses prior to breeding. Keratins are the major structural proteins of epidermis. They belong to the IF superfamily, composing extensive 10 nm cytoskeletal frameworks in cells. As basal cells terminally differentiate, they switch from expression of keratins KS & K14, to K1 & K10. Palmar/plantar epidermis expresses an additional suprabasal keratin, K9. We have characterized many of the human epidermal keratin genes and used molecular genetics to identify amino acids involved in 10 nm filament assembly. We showed that transgenic mice expressing IF assembly-perturbing 14 or K10 mutants exhibit phenotypes similar to the blistering skin disorders, Epidermolysis Bullosa Simplex (EBS) or Epidermolytic Hyperkeratosis (EH), respectively. Subsequently, we discovered point mutations in the K14/K5 (EBS) and K1/K10 (EH) genes of human patients, thereby revealing the genetic bases for these two autosomal dominant diseases. Additional sequence analyses of EBS and EH families are now required to ascertain whether these diseases always involve keratin mutations, and whether locations of mutations within keratin polypeptides correlate with disease severity. Functional analyses should elucidate whether the degree to which a mutation perturbs filament assembly correlates with disease severity. Similar strategies should unravel whether K9 mutations are involved in Epidermolytic Palmoplantar Keratoderma. Collectively, these studies are of fundamental importance to understanding the relation between keratin gene defects and this group of blistering human disorders. They are essential in developing new and improved diagnostic tools for these diseases. While dominant negative keratin mutations in mice led to a genetic understanding of EBS and EH, they have not allowed unequivocal assessment of keratin function and the significance of IF multiplicity. Thus, we do not know (l) whether cell degeneration and seemingly bi-nucleate cells arise from absence of an IF network or accumulation of altered cytoskeletal material, nor (2) whether IF diversity is redundant or whether it enables a cell to structurally tailor its cytoskeleton to suit its particular needs. To answer these questions, homologous recombination and ES technology are needed to knock- out, mutate and replace endogenous epidermal keratin networks in vivo. An understanding of keratin function and the significance of IF diversity is of utmost importance for elucidating the consequences of IF defects in human disease. This work will be essential to the development of new and improved methods to treat disorders of IF proteins.